Introduction CRITICAL STUDY SKORUPSKI ON SPONTANEITY, APRIORITY, AND NORMATIVE TRUTH

CRITICAL STUDY SKORUPSKI ON SPONTANEITY, APRIORITY, AND NORMATIVE TRUTH The Domain of Reasons. BY JOHN SKORUPSKI. (Oxford University Press, 2010 (hbk), 2012 (pbk). Pp.525. Price 66.00 (hbk), 27.99 (pbk).) Introduction An important enterprise in recent literature on normativity is the Reasons First Project (RFP), which seeks to analyze all normativity in terms of reasons. The Domain of Reasons is the most ambitious and penetrating defense of RFP to date, and should be its locus classicus for years to come. The book is an equally monumental contribution to the broader metaphysics and epistemology of the normative, defending a novel cognitive irrealist outlook that provides a compelling rival to outlooks in similarly bold books, such as Parfit (2011). The book s reach extends much farther: it examines topics of immediate interest to metaphysicians and philosophers of language and mind like modality, analyticity, rule-following, and concepts, and makes important contributions to first-order ethics and epistemology. And despite its tremendous size, the work hangs together beautifully, thanks in part to a historical narrative that connects and compares its undertaking to Kant s Critical project and logical positivism. The book divides into four parts, the first three developing Skorupski s version of RFP and the last expounding his cognitive irrealism. Part I mainly lays down the conceptual groundwork, discussing the general properties and logic of the three reason-relations he takes as primitive (epistemic, practical, and evaluative), the distinction between neutral and relative reasons, the general framework for analyzing normativity in terms of reasons, and the nature of reasons we possess ( warranted reasons ) and of correctly responding to reasons ( acting from reasons ). Part II is nominally about epistemic reasons, but half of it concerns topics usually treated by metaphysicians and philosophers of language viz., modality and analyticity. The rationale is that according to Skorupski, modality is itself a normative notion analyzable in terms of epistemic reasons, and pace the positivists, analyticity will do little to illuminate the a priori. The other two chapters in Part II directly concern the epistemology of the a priori and a posteriori, with the first setting the stage for a later discussion of the epistemology of normativity. Part III turns to evaluative and practical reasons, focusing in large part on concepts of morality and prudence. Part IV 1

develops Skorupski s take on the metaphysics and epistemology of the normative, but also discusses rule-following, concept possession, and self-determination. In this critical study, I ll focus on aspects of the book that intersect with epistemology and philosophy of mind. These aspects have received less attention than they deserve. Reflection on them reveals, I will argue, a tension in Skorupski s meta-normative outlook: we ought, I believe, either to resist his spontaneity-based epistemology of normativity or abandon the cognition-independence of normativity and accept constructivism (which Skorupski resists but doesn t sufficiently address in its best, Kantian form). After discussing this tension in 1, I turn in 2 to consider one rationale for adopting a spontaneitybased epistemology of the normative viz., that the receptive intuitionist alternative is unacceptable. I argue that this rationale is unconvincing, since there is a better intuitionbased epistemology that isn t touched by Skorupski s case against intuitionism ; notably, this epistemology is neither receptive in Skorupski s sense nor spontaneity-based. If we want to maintain a cognition-independent view of normative truth, we would do better to adopt this epistemology. I proceed in 3 to suggest that the intuitions that support the constraint on reasons Skorupski calls cognitive internalism and his claims about the range of reasons might recommend a fully Kantian meta-normative package. Nonetheless, I raise some worries about the reliability of these intuitions, worries echoed by ones I raise in 4 about Skorupski s discussion of epistemic reasons in Part I. I conclude, however, by recommending that he opt for the fully Kantian package. 1 Spontaneity and Cognition-Independence: A Tension Skorupski s outlook is Kantian in one way and un-kantian in another. On the one hand, his epistemology of the normative crucially invokes the notion of spontaneity, with Skorupski going so far as to claim that [s]pontaneity is the sole final source of normative knowledge (414). On the other hand, he denies that normative truths are cognitiondependent or more broadly mind-dependent, and in this way departs from Kant s line of resistance to global realism. Although Skorupski calls his outlook irrealist, this is not because he accepts the constructivism favored by other Kantians (e.g., Korsgaard and Markovits). It is rather because for Skorupski, Reality is the realm of substantial facts, which by stipulation belong to the causal order, and normative facts are not in this sense substantial. Normative facts are rather nominal facts i.e., normative propositions that are true in the minimalist way though these facts are, Skorupski agrees with non-naturalist realists like Parfit, irreducibly normative. While Skorupski helpfully illuminates the elusive notion of spontaneity on pp.406 410, I am left with questions about its epistemic role. I have two related worries, which lead to a dilemma. Firstly, I do not understand how spontaneity could be a reliable guide to truths about reason-relations unless one accepts a cognition-dependent, constructivist account of them. And if spontaneity isn t a reliable guide to normative truth but merely a source of internalist justification for normative propositions, I cannot see how it helps us 2

to understand knowledge of normative truths. Although Skorupski s account of epistemic warrant isn t couched in terms of reliability, he doesn t deny that reliability is necessary for normative knowledge; indeed, it is explicitly part of his view about a priori knowledge in general on p.161, and concerning the special case of normative knowledge on p.162 he just says that the way in which reliability is secured isn t via a receptive faculty. To avoid this worry, Skorupski could stipulate that spontaneity is itself to be understood in terms of responsiveness to cognition-independent normative reasons. But then I worry that so understood, spontaneity cannot play a fundamental role that would distinguish his picture from pictures held by non-naturalist realists like Parfit and earn it the Kantian framing it receives. After all, one plausibly cannot respond to a normative reason without having access to that normative reason as such, and this kind of access is itself plausibly understood partly in terms of normative knowledge. Accordingly, if spontaneity is grounded in reasons-responsiveness, normative knowledge cannot itself be grounded in spontaneity on pain of circularity or regress. Spontaneity must, it seems, explain reasons-responsiveness rather than be explained by it if it is to play a fundamental epistemic role but that way leads to constructivism. Hence the dilemma: either go constructivist or deny spontaneity a fundamental epistemic role. One is reminded of McDowell (1995), who worried that an epistemology based solely on spontaneity would lead to frictionless spinning in the void. Of course, if truths about reasons are not substantial, we do not literally want friction. But we also do not want spinning in the void. And why would beliefs about reasons that are really mine and reflective of my true nature phrases Skorupski treats as synonymous with spontaneous have a better shot at the mark of normative truth if truths about reasons are not themselves derivable from constitutive principles of self-determination or autonomy, as Kantian constructivists suggest? I see no answer. This is not to deny any epistemological significance to spontaneity. Skorupski is perhaps right that explanations that subvert or explain away the spontaneity of a response... also remove its epistemic value (409). But insofar as this claim is plausible, it is because spontaneous contrasts with factitious, where a factitious as against a truly spontaneous response or disposition is one that is accepted uncritically into one s thinking from others (406). Uncritical thinking in one sense of the phrase is unreliable, and so evidence that a response was factitious will be an undercutting defeater. So, if non-spontaneity implies factitiousness, non-spontaneity implies (or at least probabilifies) unreliability. But it doesn t follow that spontaneity is sufficient for reliability. What follows is that spontaneity is necessary for reliability, assuming and here one worries about the stipulative nature of spontaneity that non-spontaneity implies factitiousness. I don t see how a sufficiency claim would be plausible without constructivism. Now, there is surprisingly little discussion of constructivism in the book, given the putative origins of one of its most prominent versions viz., the Kantian constructivism of Korsgaard and Markovits. 1 Skorupski rejects constructivism on p.429, but there he 1 Skorupski denies on pp.11 12 and again on pp.484 487 that Kant was a constructivist about practical 3

treats it as of a piece with fictionalism, which is not how Kantian constructivists intend it. Skorupski seems in places not far from Kantian constructivism, writing, for example, that [r]eason relations... are irreal because they are objects not of imagination but of pure cognition. A priori truths about them are cognition s norms (429). If one added the word constitutive between cognition s and norms, one would have a Kantian constructivist proposal. One wonders, then, why Skorupski doesn t consider this familiar line. It could help to vindicate the role of spontaneity. Moreover, unless given constructivist mooring, Skorupski s metaphysical outlook is dangerously close to that of his non-naturalist realist opponents. Parfit, for example, agrees that normative facts are not substantial in Skorupski s sense. In Parfit (2011), he accepts Non-Metaphysical Cognitivism, and his gloss sounds a lot like cognitivist irrealism: There are some claims that are irreducibly normative in the reason-involving sense, and are in the strongest sense true. But these truths have no ontological implications. For such claims to be true, these reason-involving properties need not exist either as natural properties in the spatio-temporal world, or in some non-spatio-temporal part of reality (Parfit 2011, Volume 2: 486). Since Skorupski accepts the cognition-independence and irreducibility of these truths, it is hard to see why his view is less mysterious than Parfit s. It is not less mysterious just because Skorupski labels his view irrealist. Of course, perhaps it is wrong to see either view as mysterious. But if so, how has Skorupski made an advance over his non-naturalist realist opponents? Given Skorupski s inspiration, one would like to hear why he doesn t distinguish himself from Parfit by following constitutivist Kantians and making more of the claim that reason-relations are objects of pure cognition, and truths about them cognition s (constitutive) norms. While Skorupski denies that this view is Kant s, he happily departs from Kant elsewhere. And other theses Skorupski considers attractive could push us from what he describes as Kant s view to the constitutivism of some Kantians. Skorupski ascribes to Kant the Insight principle that no moral law applies to us whose validity we cannot recognize for ourselves, and in this way agrees that for Kant, our nature constructs our experience of the moral law as law as imperatival, binding in somewhat the way that it constructs our experience of things in themselves as objects in a causally integrated spatio-temporal field (486). Constitutivism about the moral law explains why this principle is true, and constitutivism about reason-relations would provide a similar grounding for the broader constraint on reasons Skorupski dubs cognitive internalism (p.73), which is just the Insight principle writ large. While Skorupski notes on p.487 that we will be stuck with subjectivism if we hold that we decide what our moral obligations are, constitutivists don t hold that view. We do not decide what the constitutive principles of autonomous agency are. So constitutivism doesn t face the concerns Skorupski raises. While constitutivism faces other problems, it is worth considering to eliminate the tension between Skorupski s epistemology and metaphysics. The tension is vivid when Skorupski is explaining why he doesn t follow constructivists: We must remember... the normativity, however, so this fact may explain why he doesn t feel pressure to discuss Korsgaard. 4

distinction between holding that practical principles have their epistemic basis in dispositions of the will (as argued in Chapter 16) and holding that they simply are the content of actual acts of will (487). Epistemology is beholden to truth. While we should deny that practical principles are the content of actual acts of will, we need some deep link between the will and truths about reasons if Skorupski s epistemology is to work. Constitutivism provides that link and is not the simple view rejected at the end of the quote. 2 Spontaneity by Elimination? The aim of the last section was not to undermine a spontaneity-based epistemology of normativity but to question its consistency with other parts of Skorupski s outlook. I now turn to consider whether Skorupski gives us sufficient reason to pursue a spontaneitybased epistemology of normativity. Skorupski s initial path to this view is negative. In Part II, he attacks two models of apriority he labels intuitionism and the no-content view. Intuitionism for Skorupski holds that a priori truths state substantial facts, and that we know these facts obtain by a receptive faculty of intuition (138). Intuitionism so defined is in stark contrast with his own view, on which a priori truths state nominal facts that are no part of the causal order, and our knowledge of which is based on pure spontaneity (139). While I agree that we should reject intuitionism as Skorupski defines it, there is a better view deserving of the label that lies in between this view and his. And if we want to preserve the cognitionindependence of normativity, we ought to prefer this view. As a warm-up, note that no popular general accounts of knowledge today are essentially receptive in Skorupski s sense. A receptive theory of knowledge in this sense would be a causal theory of knowledge along the lines of Goldman (1967), which requires an appropriate causal connection running from the fact that p to one s belief that p. Such theories are unpopular in contemporary epistemology; Goldman himself swiftly abandoned this view. While there are popular theories with causal elements, these elements do not involve essential appeal to a causal connection running from the fact that p to one s belief that p, which is the kind of causal connection most worth calling receptive. Consider, for example, Sosa (2007a) s account of knowledge as apt belief. S aptly believes iff the correctness of S s belief manifests some cognitive competence. While the notion of manifestation is plausibly causal, 2 it doesn t follow that on this account the fact that p must itself cause S s belief that p. Rather, the correctness of S s belief must owe in the right way to cognitive competence. The explanatory relation here runs from the competence to the correctness of the belief, not from the belief s truth-maker to the belief. There is no reason why this account of knowledge cannot apply to the a priori even if a priori truths state merely nominal facts. It is compatible with an irrealist (in Skorupski s sense) and indeed constructivist metaphysics of the normative. Perhaps some 2 Though Sosa (2015) takes the notion as primitive and rejects any simple causal analysis. 5

epistemic competences (e.g., perceptual ones) are ones whose apt manifestation requires a Goldman-circa- 67 causal link. But then receptivity merely emerges as a local feature of the epistemology of certain domains. Observe now that not all epistemologies of the a priori that appeal to intuition understand intuition as a receptive faculty. Sosa is again exemplary. Sosa (2007a, 2007b) defends the use of intuitions, but he is a reductionist about intuitions, viewing them as epistemically evaluable attractions to assent to propositions exerted by the sheer understanding of them (where the subject matter is modally strong). Intuition is not a sui generis receptive faculty for Sosa; indeed, intuitions for him are epistemically evaluable propositional attitudes. According to Sosa, these can play an epistemic role provided that they manifest a competence...on the part of S to discriminate the true from the false reliably (enough) in some subfield of modally strong propositional contents that S understands well enough, with no reliance on introspection, perception, memory, testimony, or inference (2007a: 60). Skorupski cannot complain that the appeal to reliability implicates receptivity, since he himself appeals to reliability in his account of a priori knowledge on p.161. And since a priori truths are mind-independent on his view (just non-factual), he fares no better in explaining reliability. This view is compatible with a wide range of metaphysical views. While it is consistent with non-naturalist realism, it is also consistent with views that are cognitivist and irrealist in Skorupski s sense, like the many varieties of constructivism and Skorupski s own view. One could develop a variant of Skorupski s epistemology that is a version of this view, on which a priori knowledge is knowledge grounded in spontaneous attractions to assent that manifest reasons-sensitive competences. These attractions are also reliable guides to synthetic a priori truths thanks to the fact that all synthetic a priori truths are either reason-involving truths or spinoffs of them, as Skorupski suggests in Chapter 8. I am, however, unsure that we should build the idea that all synthetic a priori truths are either normative truths or spinoffs of them into our epistemology. Skorupski s spinoff story seems defensible for simple synthetic a priori truths like the truth that no object can be red and green all over. Perhaps this truth is a priori because it is knowable by suppositional reasoning about arbitrary objects using the norm (N) on p.155. But there are hairier synthetic a priori truths, like significant philosophical truths. One would like to hear about Skorupski s philosophy of philosophy, since it is not clear that his story in Chapter 8 plausibly generalizes. Perhaps it does, but without seeing the details I would, other things being equal, prefer a simpler epistemology along the lines of Sosa s. If we opt for that simpler epistemology, we can hold onto the cognition-independence of normativity. But if we accept this package of views, the result is not a distinctive metanormative package that can claim a Kantian heritage. The meta-normative package will rather be close to Parfit s, as long as we notice that Parfit, like Skorupski, also denies that cognition-independent normative truth requires backing by a spooky ontology. 6

3 Cognitive Internalism and the Range of Reasons There are other reasons for preferring the Kantian package than that only it can provide a credible epistemology of normativity. A further reason emerges from reflection on Skorupski s discussion of what determines the range of a reason (i.e., the set of actors to whom it applies) and cognitive internalism. If this reason is good, however, it should lead us to accept a full-blown Kantian constructivism, not Skorupski s midway position. In Chapter 3, Skorupski considers three possible constraints on the range of reasons: the actors over whom a reason ranges must be ones (i) who can entertain the thought that the relevant, reason-giving fact obtains, (ii) who can entertain the content of the act for which the fact is a reason, or (iii) who can recognize de se that the fact is a reason for the act. Of these three, Skorupski regards (iii) as the most plausible, though he is uncertain about whether we should accept the cognitive internalism about reasons that (iii) induces. But if cognitive internalism is true, it may provide an abductive argument for Kantian constructivism. Plausibly, Kantian constructivism explains why (iii) is a constraint on reasons, because it holds that to be a reason just is to be a consideration that any rational agent could recognize as such in virtue of her constitution as a rational agent. And I see no clear reason to expect cognitive internalism to be a constraint on reasons if we accept a cognition-independent picture of reasons. If Kantian constructivism is also the most extensionally adequate version of constructivism (which is, I think, not implausible), then if we accept cognitive internalism we should accept it. Now, matters are not quite so simple. For there are less intellectually demanding constraints than (iii) that allow us to bring more agents within the range of a reason that Skorupski didn t consider. Even beings who lack the sophisticated concept of a reason might be able to treat certain facts as reasons of certain kinds, by being competently disposed to respond in the ways that would be supported if these facts were reasons of these kinds. 3 One might instead suggest that an agent is in the range of a reason only if s/he can (a) treat the fact that gives this reason like the reason it is and (b) do so as a manifestation of a competence to treat reason-giving facts like the reasons they are. Call satisfaction of this constraint potential reason-sensitivity. Recognition of normative truths plausibly requires potential reason-sensitivity, but potential reason-sensitivity is, one might reasonably hold, a more primitive achievement. Reason-sensitivity could also ground a less intellectually demanding account of warranted reasons and the notions Skorupski wants to analyze in terms of them viz., autonomy and self-determination. Skorupski s account of warrant is highly demanding. While he claims that it is stipulative and is happy to grant that we are readier to impute reasons than to impute the concept of a reason (59), I would still prefer to exploit weaker notions. If so, however, we probably shouldn t go looking for a Kantian explanation of constraints on the range of reasons. These points bear on the moral Skorupski draws from his insightful discussion of Williams s internalism in Part III, where he suggests that Williams s best argument rests 3 For more on this idea, see Sylvan (2015). 7

on cognitive internalism. The idea is that Williams s requirement of effectiveness (i.e., that A has a reason to φ only if it is possible for A to φ for that reason) entails cognitive internalism. I think this is false for the same reasons why I suggested (iii) was too strong. The requirement of effectiveness only implies cognitive internalism if we think that acting for a normative reason requires recognizing de dicto that this fact is a normative reason. This is an overintellectualization. Nonetheless, it is plausible that acting for a normative reason requires manifesting a competence to treat particular reasons like the reasons they are, where treating something as a reason doesn t require having the concept of a normative reason. Granting the requirement of effectiveness, the upshot is no less interesting than cognitive internalism: something is a reason for A only if A has the ability to correctly respond to this reason. But at this point, some further discussion of the distinction between normative reasons we have ( subjective normative reasons ) and normative reasons there are ( objective normative reasons ) would be in order. One could view cognitive internalism or my weaker spinoff of it merely as a constraint on subjective normative reasons. It would only imply interesting conclusions about objective reasons on the assumption that these are necessarily possessable i.e., only if the agents for whom they are objective reasons could have them as subjective reasons. That assumption is more fundamental than cognitive internalism or my spinoff of it. It should be addressed in its own right. Related thoughts led me to wonder why we should buy Skorupski s claims about the connection between wrongness and blame and his epistemic condition on wrongness in Part III. Skorupski claims that [m]oral wrongness turns neither on the agent s actual beliefs, nor on the facts, but on the beliefs warranted in the agent s epistemic state (297). But one could draw the subjective/objective distinction for wrongness and hold that this claim is true only of subjective wrongness, which is what goes with blameworthiness. One would then need further assumptions to recommend any cognitive internalist constraint on wrongness period (such as p.302 s Insight principle ), which Skorupski entertains in 12.4 5 and again in 15.9. There may be an argument for these further assumptions, but I suspect it would conflict with Skorupski s desire to preserve cognition-independence. 4 Epistemic Reasons and the Unity of Normativity The foregoing remarks bring me to a final set of comments. Epistemic reasons, we are told in Part I, have a special feature in that they exhibit a dependence on the actor s actual and possible knowledge ; if this weren t so, Skorupski worries, what gives me reason to believe that p obtains could just be that fact, the fact that p, unrestrictedly for any fact at all (41 2). But I do not see sufficient reason for special treatment. Following Skorupski himself in Chapter 5, we could distinguish between subjective and objective reasons and maintain that only subjective epistemic reasons exhibit the relevant feature. We could then note that they are no different from other subjective reasons in this way. While epistemologists focus on subjective reasons, there seems to be a clear notion of an 8

epistemic reason that isn t relativized to any particular person s actual or possible knowledge. Given this notion, we could maintain that any fact that p is indeed an objective reason to believe that p, but it can only be possessed as a (non-inferential) reason to believe that p by people with certain cognitive abilities. 4 Skorupski may be resistant because he is keen on pp.47 8 to distinguish epistemic reasons and indicators. Indicative evidence, after all, would provide the clearest example of an objective epistemic reason independent of anyone s actual or possible knowledge. But that thought aside, there is no obvious disanalogy between the epistemic case and the practical/evaluative case. Hence, the desire for unity inclines me to query Skorupski s case for thinking that objective reasons for belief aren t just indicators. Skorupski postpones this case to Chapter 9. The argument I see there, however, simply recommends a different route to treating epistemic and other reasons alike. There Skorupski suggests that indicative evidence is relative to our receptive powers, since (e.g.) a sound that dogs can hear but we can t isn t evidence for us (213). He concludes that bits of indicative evidence are indeed relative to certain possible epistemic fields. To the extent that I find the driving intuition compelling, I suspect there is a more general inprinciple accessibility principle steering my intuitions that both explains this restriction on indicative evidence and the intuition motivating cognitive internalism. If so, we still get parity because epistemic and practical/evaluative reasons are all relative to possible epistemic fields for an overarching reason (e.g., my spinoff of cognitive internalism). This is one way in which Skorupski s view could be more unified. There is another way. In Chapter 2, Skorupski maintains that there is an irreducible trichotomy of reason-relations viz., epistemic, practical, and evaluative. Skorupski mentions two views epistemicism and pragmatism on which there is just one basic reason-relation. I would follow Skorupski in rejecting pragmatism. But the version of epistemicism he mentions on p.54 and again on p.480 is not a view on which evaluative and practical reasons are reduced to epistemic reasons, but rather one on which the former are analyzed in terms of yet other normative notions. Notably, Skorupski does not rule out a version of epistemicism inspired by Kearns and Star (2009) on which reasons are evidence. 5 Of course, Kearns and Star do appeal to ought, so their full view may call RFP into question. But one could construct a variant on their view that makes no explicit appeal to robustly normative notions. Consider a view on which a consideration X is a reason for a response R in circumstances C iff X is an epistemic reason to believe that R-ing would meet its constitutive aim in C (e.g., truth in the case of belief). To be sure, particular constitutive aims might coincide with something robustly normative perhaps desire aims at the good but the general characterization of reasons can be neutral here, leaving this as a substantive issue. Perhaps Skorupski doesn t 4 For more on this suggestion, see Sylvan and Sosa (forthcoming). 5 In Chapter 9, Skorupski claims that [e]vidence involves the notion of inference, and in particular, a posteriori inference (211). So if there are non-inferential epistemic reasons, we would have to deny that all reasons are evidence. But it is a familiar point from Conee and Feldman (2004, 2008) that evidence can be understood in a wider way, allowing non-inferential reasons to be evidence. 9

take this view to be inconsistent with his trichotomy thesis. If so, however, he understates the case for unifying the reason-relations. For on this view, epistemic reasons are in a clear way the fundamental kind. 5 Conclusion A satisfactory defense of RFP would not only show us that we can analyze all normativity purely in terms of reasons, but also explain why we should prefer such analyses to similarly economical analyses in terms of other normative notions (e.g., goodness). There might, after all, be many equally parsimonious ways of analyzing normativity, as the cottage industry of X-first accounts of normativity strongly suggests. Skorupski is attuned to this concern in the Introduction, and a big-picture reason for preferring RFP there is the underlying, stage-setting thought that... normativity can be nothing more than a way of talking about that by which self-determining reasons-sensitive agents steer (4 5). I think this is precisely the kind of deeper rationale for which defenders of RFP should be searching. Skorupski s defense of RFP is extremely refreshing in being structured around such a rationale; to try to determine what the sole fundamental constituents of the normative are in the absence of a broader meta-normative program is a vain task. But if Skorupski s rationale succeeds, I suspect that it is only because it meshes with and indeed falls out of a Kantian constructivist metaphysics. For the central upshot of this discussion is that we ought either to pursue constructivism or deny spontaneity any fundamental role in our meta-normative outlook, opting for the modest intuitionism mentioned earlier. If the Critical project is to leave common cognition intact, we might be forced to pursue the latter option, since it is unclear whether there is an extensionally adequate Kantian constructivism though Kantian constructivism probably fares better than other constructivisms on this score. But the most distinctive aspect of Skorupski s project, what sets it starkly apart from the work of other reasons-firsters, is its Kantian flavor. To motivate and better preserve the distinctiveness of his version of RFP, I would encourage Skorupski to become more Kantian and accept the cognition-dependence of normativity suggested by cognitive internalism and the alleged role of spontaneity. References Conee, E. and Feldman, R. 2004. Evidentialism. Oxford: Oxford University Press. Conee, E. and Feldman, R. 2008. Evidence in Smith, Q. (ed.) Epistemology: New Essays. Oxford: Oxford University Press. Kearns, S. and Star, D. 2009. Reasons as Evidence. Oxford Studies in Metaethics 4: 215 42 McDowell, J. 1995. Mind and World. Cambridge MA: Harvard University Press. Parfit, D. 2011. On What Matters. Oxford: Oxford University press. Goldman, A. 1967. A Causal Theory of Knowing. Journal of Philosophy 64: 357 372. 10

Sosa, E. 2007a. Intuitions: Their Nature and Epistemic Efficacy. Grazer Philosophische Studien 74: 51 67. Sosa, E. 2007b. Apt Belief and Reflective Knowledge, Volume 1. Oxford: Oxford University Press. Sosa, E. 2015. Judgment and Agency. Oxford: Oxford University Press. Sylvan, K. 2015. What Apparent Reasons Appear to Be." Philosophical Studies 172.3 :587 606. Sylvan, K. and Sosa, E. Forthcoming. The Place of Reasons in Epistemology in Star, D. (ed.) The Oxford Handbook of Reasons and Normativity. Oxford: Oxford University Press. 11